Terroir. Tairrrr-wahrrrr. Roll those r's around your mouth as a Frenchman would. Gargle softly with them; squirt them through your teeth and from one cheek to the other, as if they were a 1990 Chambertin. Then go walking in late winter through the storied vineyards of Chambertin, in Burgundy, where great clods of thawed mud-terroir-will stick to your shoes. Scrape it off with a limestone shard you've plucked from one of the ancient crumbling walls. That, too, is terroir.
Does a vineyard's particular combination of dirt, rock, and exposure to sunlight-its terroir-leave an indelible mark on the wine? People in France say that vineyards just a stone wall apart-such as Chambertin and its neighbor Chambertin--Clos de Beze-will produce wines that taste different even if they are made in exactly the same way, because the terroir is different. The notion sounds mystical, even a little snooty to American ears. But lately science has been backing it up.
Perhaps the issue should have been resolved long ago, but the more we find out about wine, the less we seem to know. Wine is so much more complicated than orange juice or wheat or other agribusiness products. "We're entering the third millennium, and we're a long way from understanding the chemistry of wine," says Michel Feuillat, director of the Wine Research Institute at the University of Burgundy in Dijon. "When I started doing enology many years ago, people said there were 300 different components to a wine. Today we say there are 10,000." Modern analytical techniques have not only made it possible to detect even trace amounts of those components but, more remarkably, have also made it obvious that humans can smell and taste incredibly tiny quantities. The molecule responsible for giving wine its unfortunate "corked" odor, for instance, can be discerned (by discerning noses) in concentrations of less than one part per trillion.
That molecule, trichloranisol, has been extracted and identified. But the molecules that give a great red burgundy a "decadent, even raunchy . . . smell," as wine maven Robert Parker describes it, as well as "layers and layers of black and red fruits that virtually explode on the palate with a cascade of increasingly expanding textural sensations"-well, those precious essences are still mysterious. Knowledge of how a wine acquires that mysterious composition is just as sketchy. The list of influences is almost endless. There's the weather, of course: rain too close to harvest time bloats the grapes and waters the wine. So can overfertilization or a wine maker's unwillingness to thin the grape crop and reduce his yield during an abundant year. Then there's the skill of the wine maker in knowing when to harvest: for a red, the ideal moment is when both the level of sugar in the grape juice and the level of phenols in the skin-tannins that will give the wine body, anthocyanins that will give it color-are near a maximum.
Next comes the choice of techniques for fermenting the sugar into alcohol: what species of yeast, what temperature, and what method of dunking the floating mass of grape skins back into the juice, the better to extract all that color and body. And finally there are the barrels the wine is aged in: oak or not, if so what species, and how long should the cooper heat the staves? Wood heated for less than 15 minutes, say researchers at the University of Borde, adds hints of toast and vanilla to the wine; wood heated for longer adds smoky, spicy notes.
American experts tend to emphasize all these choices made by the wine maker as the deciding influences on a wine. "The Californians say, 'If we vinify well, if we have good vines, we don't see why our chardonnays should be any worse than yours,' " says Feuillat. Although the notion of terroir has begun to catch on among enologists at some American wineries, many will cheerfully throw together grapes from vast estates or entirely separate vineyards even when making ambitious wines. Those wines may turn out tasty and top-notch-no one disputes that great wines are made on both sides of the Atlantic-but to a French connoisseur they will lack something essential, namely the essence of a particular small patch of land. "Every soil brings its own flavor," says Damien Gachot, a young wine maker now tending the ancestral plots in Burgundy. "That's terroir. And the terroir influence is the most important one."
At a laboratory of the National Institute of Agronomical Research outside Angers, in the western Loire Valley, a team led by Christian Asselin and Rene Morlat have recently added scientific meat to that notion. Their method is simply to do controlled experiments.
Some years ago Asselin and Morlat planted a number of test parcels in their region with cabernet franc, the red varietal grape that does well there. They took care to plant all the parcels at the same density. Every year since, they have tended all the fields and all the vines in exactly the same way, and after harvest they have followed exactly the same procedures to make the grapes into wine. Thus the only differences between the wines from different parcels should be due to the effects of terroir.
Presenting such wines to a panel of experts in blind tastings, Asselin and Morlat have found that wines from certain parcels are consistently picked out as having certain characteristics, year after year. In fact, the terroir differences are of such magnitude that you don't have to be a trained gargle- and-spit professional to detect them. Confronted with three 1996 cabernet francs from three parcels only hundreds of yards apart, your correspondent, in the early stages of flu and not known for his nose on the best of days, was knocked back on his heels by a startlingly harsh "animal note" in one of the wines. It happened to be the best one, in the researchers' opinion, but the point is that the animalism was absent from that wine's fruitier peers. A Yoo-hoo drinker could have told the difference.
The Angers researchers have figured out how such differences can arise from differences in terroir. In their region, at least, the key variable is soil temperature: some soils warm up sooner than others in spring and stay warmer throughout the growing season. A vine with roots in such soil gets a jump start over one on a less favorable parcel. Its grapes start making sugars and phenols sooner, and they never lose their lead. The inevitable and perhaps enviable result is recognizable differences in the wine.
A thin layer of soil on a basement of porous chalk, of the kind that was used to build the Loire Valley castles, is one of the best terroirs for cabernet franc, especially when it's on a sunny slope; a thick layer of clay is one of the worst because it holds a lot of water and takes a long time to warm. But there are many possible variations of basement rock and soil type. As Asselin and Morlat map the terroirs of Anjou (the region around Angers), they often find dozens of distinct units in a single village, each capable of producing a distinct wine.
In Burgundy this news does not surprise anyone. The beginnings of a controlled terroir experiment have long been in place here: all the reds worth mentioning are pure pinot noir, by fourteenth- century ducal decree, and all the whites are chardonnay. The greatest ones, the grands crus, are named for, and by law must come exclusively from, fields that may cover just a few acres. The fermented essences of these fine plots of land often fetch upwards of a hundred dollars a bottle when they are obtainable at all. Wine makers here do not need scientists from upstart Anjou to explain to them the value of terroir. When Morlat once offered to apply his methods to the study of some of their prime parcels, the response of the vintners was layers and layers of ripe red raspberry.
"Burgundians aren't too hot on having a terroir study here like the one they did at Angers," says Feuillat. "They say, 'You're going to demystify everything. If you start saying a grand cru means such and such a percentage of clay and limestone, such and such a slope and nutrition of the vine-it will lose all its sacredness.' " And perhaps some of its worldly value as well.